Has anyone done a proper shootout between flat bed and camera scanning?

What is the verdict as far as speed and quality?

Your question is far too general.
Which flatbed scanner vs which camera/lens combination?
The permutations are endless but imho a very good camera plus a very good lens will outperform almost any flatbed scanner.

It is not good to overthink things Belinda or you will never get going…you quit before starting. Just submit what you or others have Belinda, even if not perfect.

My interest is what the average hobbyist may use. Decent equipment, but it does not have to be top end MF gear…but it could be. The bigger the range, the better. It is this camera setup versus that scanner Belinda. If you have enough people submitting tests, then you have a collection of varied results to look at. It all counts Belinda. I’m not into words and charts much, I like seeing photos for comparison Belinda.

Here are some photos from an old website I had on Tumblr comparing digital and film cameras of all sorts. In 2019 Tumblr deleted the website along with 47 other websites I made. I don’t have time or interest in recreating all the lost work. And sadly, The Wayback Machine didn’t archive much of it before it was wiped out. But here are a few examples of what it contained Belinda.

And these are only one sample image from a few cameras. I would do many tests (close, medium and far distance) of different things to compare using many more cameras.

6.1 mp Epson RD-1s full

6.1 mp Epson RD-1s crop

16mp Fuji X-E1 crop

Leica M6 35mm Kodak Ektar ISO 100 Film full

Leica M6 35mm Kodak Ektar ISO 100 Film crop

So, from my many tests Belinda, I concluded flatbed scanned negative film on an Epson scanner is= to about 3 or 4mp. But…I didn’t have a drum scanner to try and I didn’t have a decent camera scanning setup. It does not mean the tests are wrong, it only means those are the tests for that equipment. And in the big picture, this is the results the average hobbyist that does not get drums scans gets.

I’m not ‘overthinking’ anything. Tests like these might be interesting and good fun to some, but unless carried out in some kind of controlled environment using standardised targets, they are pretty meaningless.
And I haven’t quit on anything, I just don’t believe in wasting my time.


Belinda is correct about all of this. It is the kind of question which doesn’t even make sense to ask unless one is comparing specific scanner models with specific cameras. And one does need to use industry-standard resolution targets to get any meaningful answers. Those photos of the thermometer give about as much usable information as you’d get testing whether a truck can drive along a four lane highway.

1 Like

OK, that is fine. But what tests do you have to offer to replace the ones I’ve submitted?

Now, I take it you have a camera scanning setup. So, scan with your best camera or a variety of cameras and scan with your best or a variety of flatbed scanners then submit the results…very, very simple.

You know what I’ve submitted is wrong…so let’s see what you can submit that is right Mark and Belinda. I’m always looking for the truth, so let’s see the truth.

Sure, you are Belinda. Overthinking things like hell. You suffer from perfection dis-ease. Those type of people get stuck if they can’t do things perfectly.

As I told Mark, you know my tests are wrong Belinda, so you must know what is right. So, what have you have done so far Belinda with comparison testing?

This is not some trick question.

I don’t have camera scanning setup to speak of. Even if I do get one, 42mp is the max camera I have. I know Mark has a higher MP camera and has tested various lenses. Also, some other members must have done tests. I’ve been a street photographer for 50+ years, so I’m used to having things imperfect…but I do the best I can with what I got.

…so where are these comparison tests?

I’ll send mine in someday if and when I get a camera scan setup. I’m not shy about it. I just don’t have the camera scanning gear. All I got is a variety flatbed scanners.

Now… let me make one thing clear.

The tests with the thermometer were not meant to replace flatbed scanner versus camera scanner tests. The tests were sent in just to show Belinda not to get stuck with her endless permutations dead end. You just test what you got. It all counts Belinda. I just want to know how camera scanning compared to flatbed scanning Belinda. Is that too hard to test? Or will the endless permutations preclude any testing to ever be done Belinda?

This is a big forum…no one has ever tested the comparison??

…then why care for a proper shootout? :wink:

Have a look at this video. It’s long, but I think it’s comparing the different ways in fairly proper manner. Beware: don’t let the “fun” parts fool you.

You are right up to a point. Perfection can be the enemy of the good and I completely admit that I have perfectionist tendencies. That said, I return to what I said before: if you want to have some fun making largely meaningless comparisons, go ahead, knock yourself out, especially if, as you say, you are used to having things imperfect.

I’ve done my own testing and research to arrive at a set up which gives me very good/excellent results by any parameter. For 35mm, a Nikon Z7 and bellows mounted Minolta scanner lens with flash. For 6x6, a column mounted Nikon Z7 with an adapted Olympus 80mm F4 macro bellows lens with a diffuse colour correct light source. My aim would be to replace this with a Printing Nikkor in due course. This set up isn’t as sophisticated as some of the quite stunning rigs I have seen illustrated on this community website but it is accurate and I’m only dealing with a small number of scans.

So, I am not criticising you or saying you are ‘wrong’, I’m just saying that the only test which counts is the one done to evaluate one’s own equipment against a known standard. Anything else may be amusing but is largely pointless IMHO. I won’t be wasting my time on shootouts but don’t let me hold you back if that’s your thing.

Thanks for the link. A long and interesting video but ultimately an invalid result. He may be a wiz at flat bed scanning but has no idea how to do a good camera scan. No wonder it came out worst.

What is it that he should have done instead? What gear and techniques should he have used instead? Please share your knowledge so that we can all progress. Thanks.

At 12:20 into the video, Nick tells us how he camera-scans his negative.
Apart from using f/11, things sound good enough imo.

From your comments here, you are very knowledgable and have no doubt watched the blog which you linked. My principal points of criticism are as follows:

  1. A 20Mp camera is not the fairest starting point in a shoot out with a drum scanner as regards resolution, dynamic range or noise.

  2. The Canon 100mm macro is not a stellar performer being rated the worst performer in the parameters that matter at 1:1 on closeuphotography.com. Also, as you say, shooting at F11 doesn’t help either. Remember we are up against a drum scanner here.

  3. Shooting directly down at a light table with no attempt to mask off the negative from stray light is a schoolboy error and a recipe for all kinds of trouble with flare, reflections, light bleed, strange artefacts and loss of sharpness.

  4. Using a tripod cranked out over a table is a recipe for shake and back ache - which the tester complained about and which he cited as a drawback of his camera scanning workflow.

  5. We don’t know the CRI of the light table - really important if you are to avoid colour casts and difficulties in colour balancing related to spikey spectral responses from low CRI light sources.

  6. Stitching 12 images by manually moving the camera and/or the light table is hopeless. At macro magnifications individual frames should be advanced exactly in line horizontally and vertically or they will never merge perfectly. A geared rail is necessary for this but there is no evidence of any kind of precise method of moving the light table or camera being used. Pixel shift with fewer frames is also an option.

When you compound all the errors in the camera scanning part of this exercise it is little wonder that the tester had so much difficulty getting a decent inversion in Negative Lab Pro. Rubbish in, rubbish out.

While I applaud his skills in flat bed scanning, this shootout amounted to little more than an exercise in confirmation bias.

A rig for a fairer comparison would have been:

  1. A current, high megapixel camera.

  2. A top quality macro lens or better still, a good scanner lens.

  3. A rigid copy stand to support the above.

  4. Having the negative properly masked off from any stray light.

  5. A colour correct (high CRI) light source.

  6. A geared mechanism for advancing frames across the film in an accurate manner.

Had that been done, I believe the results would have been very different.

When I consider Nick’s source material and expectations or requirements…

  • I hear him say that the scanner is least sharp and that he does not care that much for sharpness, but for colour and ease of use
    a) I notice that he simply does not like the “Canon” colours he gets out of his camera scans and that he seems to be unable to change these colours due to lack of patience and skills
    b) I understand his preference for scanner scanning and the resources provided by negafix and the lack of need for stitching

Nick’s want for specific colour expression and low interference workflow must lead him to his conclusion and I agree with it under said conditions.

For those who are willing to camera scan and use stitching, technique and gear should follow what I herewith label “Belinda Guidelines”. They should provide the grounds for optimal input material for a digitizing process that will still take quite some effort.

The question remains: how far is one willing to go under the 80/20 rule?
I assume that the answer will always be highly personal one.

I agree.
Nick clearly has specific requirements and has evolved a specific workflow that satisfy those. He uses a flatbed scanner because it quickly and easily gives him the colours he likes and is not so concerned with sharpness. Fair enough…BUT his test was to compare three scanning methods with regard to sharpness, colour and workflow.
He has evolved a flatbed scanner workflow that is efficient though by no means quick or straightforward (inverse wet mount) with which he has become very familiar through constant practice.
Drum scanning involves no skill at all on his part and only requires inversion by him. His colours were 99% to his liking and with a little extra effort would probably have got there in the end.
His attempt at camera scanning was as a complete newbie with inadequate equipment and a total unfamiliarity with the workflow. Unsurprisingly, he struggled at every stage and got the result he deserved. This is more a reflection on his execution of the test than on camera scanning per se which, if it had been carried out with the same preparation and care as his flatbed scans, would certainly have produced a different result.
The comparison was unfair and the results invalid.
In answer to your final comment, any method of scanning requires the right equipment and good technique and the effort needed becomes less with familiarity and practice.
How far one is willing to go under the 80/20 rule is of course subjective but also governed by practical constraints. A commercial photographer doing a large number of scans has to balance the time spent against his fees and profit margin. The fine art photographer or hobbyist often doing scans one at a time, has no such constraints - it’s a labour of love. You pays yer money and you takes yer choice.

Well, a comparison gives you feedback as to the abilities of your gear or setup. But why should have to explain this to this forum??

I hope there are some GD scanner vs flatbed tests in this thread. This is the first post I looked at. I let this thread stew for some time. If I don’t see real some tests in photos, I’m not going to waste time here. I’m not here to be Belinda’s pen pal, I need some real feedback and not BS.

I don’t use the conversion software sold here. I have Lightroom 5 and it won’t work with it. I asked a stand alone version of the software me made, but I’m only interested in it for converting neg to pos and not post work. Lightroom is sufficient for my work. So, camera scanning is my interest and not the software here.

Now, I NEED to do camera scanning, but what is all of your excuses?

My need is the archival material I deal with is not conducive to flatbed scanning. Why on earth would anyone want to camera scan if they can’t get real feedback because of permutations?

Do you do camera scanning for workflow?

OK, but for workflow you may be getting subpar results. You won’t know if you can’t or won’t test…because of permutations or a closed mind.

For instance…

I shoot a tons of low light night photos. I test all my lenses wide open for sharpness, as well as the next 4 stops. Why use a subpar lens that you know has poor sharpness wide open if you think you will be shooting wide open?

For this photo I used a lens that shoots good wide open. I shot it shot at f1.4.

Little Dickey 2016

My lenses may not be the sharpest on the planet wide open. That is not my concern. My concern is that I use the sharpest lens I have available to me for wide open shooting.

For IR flash I experimented for 4.5 years perfecting it. Lots and lots of Belinda’s permutations involved.

Testing is what perfected IR flash for me, not charts and graphs.

Transwoman and Friend 2015 IR Flash Candid

Selection from Piercing Darkness 2015 IR flash Candid

Well, I will see what ‘developed’ here. If no photo tests have been posted, I’m out of here. And if so, thanks for everyone’s participation and sorry things didn’t work out. The forum looked very promising, but looks only go so far…you gotta test things out.

Sorry but I have no idea what you are rambling about.

Nick Carver’s verdict is in his video. Each test depends on what is tested and how. If you do things differently than Nick, your conclusion might be different. If you have a scanner and a camera, you can test and compare results that apply to your own situation, which is by far the best thing you can do.

“Each test depends on what is tested and how.”
If you do one part of a test well and another part of the test badly, (as Nick did), the ‘verdict’ is a meaningless nonsense. And that is my point. It’s like testing two racehorses - one sound, the other with a broken leg - and then declaring the sound horse the winner:)

You seem to be assuming that I don’t test anything. I do indeed test - which is why I sold my Nikon 9000ED scanner and now use a high Mp camera in combination with a superachromat scanner lens. I don’t however make testing my hobby.

I haven’t been back here for a couple of weeks and I didn’t receive any email notifications until this evening, so I hadn’t seen the subsequent discussion. Let me say first that I’ve been scanning for about 20 years and “camera-Scanning” for about 8 years in both cases with various hardware and software. I wrote the e-textbook on SilverFast 8 and related workflows with Lightroom and Photoshop, available in PhotoPXL.com in the store; I wrote a reference article covering a number of scanners for Lumionous-Landscape.com and recently wrote an up-dated camera scanning article on PhotoPXL.com, so this isn’t a subject I am new to and don’t need a heap of detailed testing to know what’s better than what. You can read my articles and take away what you want.

My response to Daniel consists only of several elements,
(1) Let us be clear about what we are testing for, or put otherwise what interests us technically speaking. I am primarily interested in resolution, tonality and control over tonality. Daniel, you may have other priorities so we can easily talk at cross-purposes. Now you know mine.
(2) My camera-scanning set-up out-resolves every flat-bed or film scanner I’ve been familiar with (you can visit the menu in the LuLa article). That observation emerges from multiple tests with resolution targets, and there is no doubt about it. The only issue is that it probably out-resolves the media being digitized.
(3) I have far more control over tonality, particularly for highlight and shadow detail at the capture stage with a camera, because I can overcome dynamic range limitations of both scanners and cameras by doing multi-exposures; using a camera for this purpose and then blending the exposures in Lightroom provides far more flexibility and effectiveness than available from any scanning software on the market. Of that I have no doubt because I have done it with both technologies.
(4) Both SilverFast and Vuescan have effective tools for converting negatives to positives and achieving good colour that then requires custom adjustments. Negative Lab Pro goes about this in a different way specifically designed for raw camera captures and Lightroom. It produces good colour that then requires custom adjustments.
(5) So much for the generalities. The comparative quality of what emerges from either technology depends on so many factors including one’s ability to use them properly that one needs to be very specific about exactly what is being compared with what and how it’s being done to know which is better in what respects. There is a real limit to the generalities. The variety of solutions out there are vast, so one needs really to be very selective and specific. So this is not a simple question.
(6) A comment on lenses and f/stops: most macro lenses are NOT specifically made for flat-field copy work, which is what we need for photographing film media. Ideally, one should use a bespoke lens for photographing flat media. And with such lenses, the general recommendation is that they perform optimally wide-open.


@Mark_Segal, the link to LuLa points into the void. Maybe you can post the article differently?